This is a proposal to develop and refine 1H NMR spectroscopy and imaging technology for the prediction and early detection of the response of non-Hodgkin's lymphoma (NHL) to chemotherapy, immunotherapy and radiation therapy. The study will be performed on a xenograft model of human diffuse large cell lymphoma WSU-DLCL2 in SCID mice. This model will be used to develop and test pulse sequences, evaluate their sensitivity for detection of response to chemotherapy, immunotherapy and radiation therapy and to examine the biochemical and physiological mechanisms underlying therapy induced spectroscopic and imaging changes. A selective multi-quantum coherence transfer pulse sequence, SeI-MQC, for detection of lactate (Lac) and total choline (TCho) will be hybridized with a Hadamard slice selection pulse sequence, and software for implementation of the hybrid Had-SeI-MQC sequence on Varian ANOVA will be written. Standard MRI pulse sequences for implementing dynamic contrast enhanced (DCE), diffusion weighted imaging (DWI), T2 weighted imaging (T2WI) and Tip_weighted imaging (Tlp) on the Varian will be streamlined so that a complete examination consisting of MS-BASSALE or Had-SeI-MQC combined with DCE, DWI, T2WI and Tlp (or a subset of these imaging sequences) can be implemented on the NHL model within one hour. This animal study will serve as the basis for clinical implementation of this protocol that will be submitted as a future proposal to the NIH. This program builds on previous research on murine tumor models which has demonstrated that decreases in the Lac resonance of tumors provide an early and sensitive indicator of tumor response to chemotherapy and radiation therapy. In addition, a multicenter clinical 31P MRS study of NHL patients in which this laboratory is participating has demonstrated that the (PC+PE)/NTP ratio measured prior to initiation of therapy together with the International Prognostic Index (a clinically based assessment index for NHL) can predict which tumors will exhibit a complete clinical response (CR) as opposed to a partial response (PR) or progressive disease (PD). It is hypothesized that 1H MRS measurements of TCho will provide the same prognostic information at much higher spatial/temporal resolution and that inclusion of response sensitive MRI data will further improve on the predictive capacity of this method. However, the present study of the animal model of NHL will only evaluate the capability of these methods to detect therapeutic response at an early stage.